Skip to main content

Full text of "Physico-chemical Evolu.of The Sambhar Lake System Rajasthan"

See other formats


A thesis submitted 

in partial fulflllmera of the requirements 
for the degree of 




to the 


December, 1996 


central Ll 8 RA »> 

I I- T., 

CE-i??6’H-fAM> PHy 


t.I.T. F.:4Rpur. 


Certified that the work presented in this thesis entitled “PHYSICO-CHEMICAL 
carried out by Sri Saroj Kumar Panda (Roll. No. 9510337) under my supervision and 
has not been submitted elsewhere for a degree. 

Rajiv Sinha 
Assistant Professor 
Deptt. of Civil Engineering 
Indian Institute of Technology 
Kanpur-208 016 


The author feels sincere pleasure in expressing his most sincere gratitude to Dr. 
Rajiv Sinha of Civil Engineering Department for his valuable guidance and constant 
encouragement through out the course of this work. His clear deep insight into many 
problems encountered during the whole period of this work. The author is also grateful to 
Dr. B. C. Raymahashaya, Department of Civil Engineering, IIT Kanpur and Dr. R. P. 
Mathur, Department of Civil Engineering, TIET, Patiala for several discussion and 
suggestions during the first phase of field work. I am also thankful to Dr. R. P. Singh for 
his timely help. 

The author is indebted to Mr. Dadheech (G.M.), Sambhar Salt Ltd. and his whole 
organisation for providing field guidance and lab. facilities. 

The author acknowledges the help received fi’om Asim Chatteijee, Vikrant Jain, 
Pradeep Sahoo, Umesh Singh and all members of Engineering Geology laboratory in 
doing this thesis. 








1.1 General 1 

1 .2 The Sambhar Lake And Its Surroundings 3 

1 .3 Review Of The Earlier Work Done 5 

1 .4 Objectives Of The Present Work 7 

1.5 Plan Of Work 8 

1 .6 Organisation Of Chapters 9 


2.1 Introduction 10 

2.2 Regional Stratigraphy And Structural Evolution 10 

2.3 Geomorphology, Drainage And Sedimentation History 14 

2.4 Neotectonics 17 


3.1 Introduction 19 

3.2 Data Available And Methodology 21 

3.3 Digital Image Processing Of Satellite Data Of Sambhar 

Lake Area 23 

3.3.1 Spectral Enhancement 24 

3.3.2 Spatial Enhanement 24 

3.3.3 Multiband Enhancement 26 


3.3.4 Band Ratioing 3 1 

3.3.5 Classification 32 

3 .4 Data Analysis And Interpretation 36 

3 .4. 1 Fluvial Processe And Landforms 38 

3 .4.2 Aeolian Processes And Land Forms 42 

3 .4.3 Lacustrine Processes And Landforms 44 

3.4.4 Geomorphology And T ectonics 47 

3.5 Summary 48 


4.1 Introduction 50 

4.2 Investigation procedure 51 

4.3 Mineral analysis of the rock samples 5 1 

4.3.1 Methodology 51 

4.3.2 Results and interpretation 52 

4.4 Grain size analysis of sediments 58 

4.4.1 Methodology 58 

4.4.2 Results and discussion 58 

4.5 Clay mineralogy analysis of sediments using XRD 64 

4.5.1 Methodology 64 

4.5.2 Results and interpretation 65 

4.6 Summary 74 


5. 1 Physical evolution of Sambhar lake basin 75 

5.2 Geochemical evolution of sambhar lake 78 

5 . 3 Suggestions for future work 82 






Fig 1 . 1 Location map showing study area and sampling sites 3 -a 

Fig 2.1 Map showing the distribution of the pre-cambrian crystalline rocks 

on the Aravallimountain belts and its surounding plains 1 1 

Fig 2.2 Neotectonic map of the north-eastern Rajasthan 16 

Fig 3 . 1 Typical reflectance curves for selected common natural objects 20 
Fig 3.2 Conceptual diagram showing various parameters with their 

sequence in nature 37 

Fig 3.3 Geomorphological map prepared from digitally processed satallite 

images aided by toposheets and ground check 39 

Fig 3.4 Structural map of the study area prepared fi'om filtered (high pass 

kernel) satellite images of the study area 40 

Fig 4. 1 Bivarite plot of the sediment samples of the Sambhar area 60 

Fig. 4.2(a) Log-normal and fi'equency curves of the sediment samples fi’om 

the Mendha river channel sediments 61 

Fig. 4.2(b) Log-normal and frequency curves of the sediment samples fi’om 

the lake bed at Jhapok 66 

Fig. 4.3 X-ray chart of rocks, sediments and soil collected in the field 68 

Fig. 4.4 X-ray chart showing depth wise clay mineral association of 

sediments of Gudha 68 

Fig. 4.5 Block diagram showing depth wise variation in concentration of 
different mineral 69 

Fig. 4.6 X-ray chart showing depth wise clay mineral association of 

sediments of Gudha 70 

Fig. 5.1 Tectonic model showing diffemt stages of Sambhar lake basin 
evolution 76 

Fig. 5.2 Schematic diagram showing geochemical evolution of the Sambhar 
lake basin 79 



tables page No. 

Table 2. LLithostratigraphic framework of the Aravalli Mountain Range 11 

Table 3.1 Specification of the remote sensing data used 21 

Table 3.2 IRS spectral bands and their applications 22 

Table 3.3 ; Covariance Matrix of the Principal Components 29 

Table 4. 1 : Statistical grain size parameters of different samples 59 

(J = Jhapok, M = Mendha) 




Plate 3. 1 Band 3 image of the Sambhar lake and the adjoining area after 

contrast stretching 27 

Plate 3.2 Band 4 image of the Sambhar lake and the adjoining area after 

contrast stretching 27 

Plate 3.3 Spatially enhanced image of band 4,2,1 using edge enhancing high 
frequency kernel 27 

Plate 3.4 Multiband enhancement — standard FCC prepared by coding 

band 4,2 and 1 data on red, green and blue colour respectively 27 
Plate 3.5 Multiband enhancement — FCC prepared by coding band 4,3 and 

1 data on red, green and blue colour respectively 30 

Plate 3.6 Principal Component Transformation (PCX) — enhanced colour 
composite by loading the images (PCI, PC2 and PC3) on planes 

red,green and blue respectively 30 

Plate 3 .7 Band ratioing — band ratioed image using band 4 and band 3 30 

Plate 3.7 Band ratioing — band ratioed image using band 1 and band 2 30 

Plate 3.9 Classified image using statistical clustering of pkels of Pcs 35 

Plate 3.10 ISODATA classified image showing various classes in lake 

water 35 

Plate 3.11 Standard FCC showing alluvial fans around Aravalli hills 43 


Plate 3. 12 Ground photograph showing alluvial fans and aeolian sand 

deposits near lake bed, Nawa 43 

Plate 3.13 Ground photograph showing barchan type sand dunes developed 
on Mendha river channel 46 

Plate 3.14 Standard FCC showing different parts of the Sambhar lake 46 

Plate 3.15 Ground photograph showing a part of an entrenched stream 

with its thick soil embankment suggesting uplifhnent of Aravalli 
hill at the background 48-a 

Plate 4. 1 (a and b) Photomicrograph showing the pressence of mica in the 

rock (450X, crossed nicol) 53 

Plate 4.2 (a and b) Sericitisation along the feldspar edge 

(450X, crossednicol) 54 

Plate 4.3 Plagioclase feldspar (Albite) showing twinning 

(45 OX, crossed nicol) 55 

Plate 4.4 Doleritic rock showing plagioclase laths surrounded by chloritised 
augite(450X, crossed nicol) 55 

Plate 4.5 Quartzites at lake bed showing deposits of iron oxides in 

them(450X, crossed nicol) 57 



The Sambhar lake, spreading over three districts, namely Jaipur, Ajmer and 
Nagaur of Rajasthan forms the eastern fringe of the Great Indian desert and is bounded by 
the neotectonically active Aravalli range. The region around this lake has varying thickness 
of alluvial and aeolian sand, kankar, calcrete and silcrete deposited on an uneven basement 
of Banded gneissic complex in the form of sand plain. An attempt has been made to study 
the geomorphological/neotectonic details of the area and to classify the region in various 
zones on the basis of neotectonism of the area. On the basis of sedimentology and clay 
mineralogy the source and occurrence of salt in Sambhar lake brine has been explored. 

The digital processing of IRS LISS2 data has enabled the delineation of various 
cover units and geological structures, by enhancing the subtle tonal variations. Different 
image processing techniques such as false colour compositing, contrast stretching, edge 
detection , band ratioing and principal component analysis have been used to enhance the 
lithological and geomorphological features and their effect on the associated features like 
vegetation and soil. Several tectonic features have been identified with the help of 
geomorphologic units like drainage pattern, paleochannels, etc. and the region has been 
classified in terms of four tectonic blocks. 

Using grain size and clay mineral analysis of sediments in the lake and comparing 
their characteristic properties with rock attributes of the surrounding Aravalli hills and the 
various parameters of the surrounding alluvial and aeolian plain, an attempt has been made 
to describe the source and occurrence of salt in Sambhar. 




The first scientific study of lake, in about 1869, dealt mainly with the physico- 
chemical relations of lake Geneva, Switzerland. Since then, a variety of aspects related to 
lakes have been looked into. However, very little attention has centered on the geological 
aspects of lakes. All lakes are transitory in the geologic record because the very nature of 
the lake basin, a topographic low completely surrounded by higher areas, insures its 
inevitable destruction. Thus, many lake basins, often during the short life span of a man, 
pass through a recognisable cycle of destruction, from lake to pond, marsh, swamp or to 
dry land (Reeves, 1968). 

The chemical composition of any lake water will ultimately depend on those 
elements and compounds in solution, suspension, and/or those accumulating along lake 
bottom, most of which come secondarily to the lakes by runoff. Other substances are 
produced by chemical reactions between different elements and/or compounds or as by 
organic activities (Eugster and Hardie, 1978). 

Saline lakes are hydrologically closed basins which form in areas where 
evaporation exceeds precipitation (inflow). The total inflow is however is sufficient 
enough to maintain a standing body of water. Chemically, saline lakes are rich in Cf, Mg^^, 
Na^ and S 04 ^’ ions which yield, in equilibrium in a saturated solution, only five solid 

phases i,e. carbonates, sulfates, chlorides, borates and nitrates (Reeves, 1968). Beadle 
(1974) discussed the possible criteria to distinguish fresh water from saline waters and has 
chosen the boundary at about 57oo dissolved solids, based principally on biological 
tolerance (Eugster and Hardie, 1978). Saline lakes may contain a variety of salt-tolerance 
plants from blue-green algea to sedges. 

Remnant of pluvial lakes occupy closed basins in arid regions characterised by 
internal drainage. In such climatic regions the rivers, if present, never reach the sea. These 
are termed endorheic areas because they are under-drained by streams. The realm of 
desert limnology, by definition, coincides with re^ons where closed basins are found 
(Cole, 1979). Sambhar, Didwana and some other lakes of Rajasthan are characterised by 
endorheic drainage, with neglisible depth, containing only quaternary deposits (Biswas, 
Chattopadhaya and Sinha, 1982). 

Saline lakes differ from fresh water lakes in two respects; (a) they continually 
e5q)erience a change in chemical composition and (b) there is a continuous increase in 
concentration of various salts in lake water, mainly due to a high evaporation rate. 
However, the increase in concentration of salt in a closed basin is not infinite. As water 
flows into the dry lake basin, perhaps with the first rain of the late summer or early fall, it 
carries the available salts in the catchment rocks and also dissolves the surface salts of the 
lake provided they have not been removed by deflation (Reeves, 1968). Even so , this is 
the time when the waters of the closed basins are the freshest. As precipitation declines, 
generally after the early spring rains , evaporation becomes greater than the inflow after 
which then starts the inevitable process of concentration of the salts in solution. 



Spreading over three districts of R^asthan, namely Jaipur, Ajmer and Nagaur, 
Sambhar is the largest inland saline lake of India covering approximately 225 sq. km. of 
area. Located in a gap of the Aravalli range, about 80 km. west of the pink city, the 
Sambhar lake is a hollow in a vast stretch of almost flat sand body. The present study is 
centered around the Sambhar lake (Fig 1.1) falling between latitude 26®52^N-27®02^N and 

longitude 74®54^-75®14^E covered in four Survey of India toposheets nos. 45 - 7 , 45^, 

16 4 

J N 

45— and 45 — . Sambhar is a shallow lake, reaching only about 3 m at its deepest (Gopal 

& Sharma, 1994), with an average depth not exceeding 0.61 m. The maximum length of 
the lake basin is 22.5 km, while the width ranges from 3.2 km. to 11.2 km. The lake bed 
(360 m above sea level) is almost flat, with a slope of less than 10 cm per km. ( Gopal & 
Sharma, 1994). 

Climatically, the study area is situated over a transitional area with arid climate at 
the west and semiarid climatic zone towards the east of it. The total climatic factor of the 
area is governed mainly by monsoon and the physiography of the area , i.e. the Aravalli 
range. The southwest monsoon that originates in the western part of the Indian Ocean as 
the South east trade wind, is drawn north of the equator by a low pressure area over 
north-west India. As it crosses peninsular India, part of it swings north and then west over 
the Ganges plain, to lose the last of its moisture on the eastern slope of the Aravalli range. 
The average annual precipitation over this region (the study area) is 550 mm to 650 mm 


Fig. 1.1 Location map showing study area and sampling sites 

and sometimes up to 800 mm. Another cause of low annual precipitation around this area 
is the relatively warm dry, anticyclone air that overlies the near surface air of the 
southwest monsoon (Pramanik & Ramanathan, 1952). This absorbs any moist air that rises 
along Aravalli range and results in the dissipation of any cloud and decrease the likely load 
of rain over this region. The southwest monsoon is though replaced on the autumn by light 
winds of northeast monsoon, but having had a continental route before reaching this area. 
The annual rainfall intensity as estimated by the meteorological department is 17.5 mm 
(Sharma, 1990). The average annual temperature of this region is 23®C, the maximum 
being 45°C. All these climate data nevertheless, suggest that the study area belongs to a 
semi arid region. 

Countless effort is needed to prepare a proper note elaborating the importance of 
the study area. This area is equally important for scientific research work and economy of 
the country. Within the scientific community the geographers are turned on by this typical 
piece of transitional climatic zone sandwiched between arid and semiarid climatic zones. 
The Sambhar lake has drawn considerable attention firom the ecologists as a natural 
laboratory to study the ecological diversity. It has the distinction of a “wetland of 
international importance” under Ramsar convention (1971). The problem of algal growth 
in the salt pans and reservoirs of the Sambhar lake is providing a tool for interdisciplinary 
study like biogeochemistry. The geological interest in to Sambhar lake has been strongly 
foccussed on the origin of hypersalinity in lake water. 

The commercial utilisation of natural resources i.e. salt of Sambhar lake marks the 
industrial importance of the area. The Sambhar lake produces over 2 million tons of salt 


per annum. In a major attempt to revive the Sambhar Salt Limited the Rajasthan govt, 
recently opened a large part of the lake for private organisations for a better exploitation 
of the salt resources. The adverse impact of industrial growth in the area on the delicate 
ecological system of the lake has also been a matter of extensive debate among the 
ecologist and the environmentalists. 


A good number of research papers exist on the origin of hyper salinity in the 
Sambhar lake. A number of theories exist to explain the origin of salt in Sambhar lake, but 
non of them are satisfactory and there is still a lack of consensus among the research 
workers. Nevertheless, the existing theories on the chemical evolution of Sambhar lake 
have helped in adopting an integrated approach for the present study. 

The earliest study on the salinity problem in Sambhar lake dates back to 1909 
when Holland & Christie floated the wind borne theory, arguing the possibility of wind 
transportation of salt fi’om the Rann of Cutch. But, the physiographic setting i.e. the lake 
bed at 360 m above mean sea level, and the wind mechanism of the area does not fit to 
their theory. Godebole (1952) postulated the sea water origin theory citing that in the 
geological past the Tethys sea extended up to the Aravallis and the valley now occupied 
by Sambhar lake was, at one time, a part of the ocean. The lake is thus a remnant of the 
sea. After the retreat of the Tethys sea, the lake got silted up and dried, leaving the 
accumulated salt in its sediments. Recently, Ramesh et al. (1993) have refuted the marine 
origin on the basis of limited isotopic analysis of water firom Didwana, Sambhar and 


Kuchaman lakes of Rajasthan. Biswas et al. (1987) have reviwed the existing theories and 
admitted that the enwronmental factors of extreme aridity and evaporation coupled with 
centripetal dranage may be the prime cause of input of salts into the lake. Bhatacharya et 
al. (1982) found a striking similarity in the chemical composition of the well brines and silt 
samples from Sambhar and Didwana, and thus, pointed towards the probability of a similar 
origin of saUnity in the lake. According to them, the thick beds of halite, as recently 
discovered in Didwana, may also be the source of salinity in the Sambhar. 

In a recent study, on the basis of Na/Cl and C1/S04 ratios along Avith isotopic data, 
Yadav (1995) discarded the earher theories regarding the chemical evolution of the 
Sambhar lake water. After a thorough isotopic study of the Sambhar lake water, rain 
water and the ground water he concluded that the salt in the lake water is probably the 
result of evaporation of multiple component of source water viz. river water, ground 
water and atmospheric precipitation. In an attempt to fit his data to Eugster-Hardie model, 
he assumed that there is no interaction between solid formed and the residual water during 
the evaporation and also neglected the “ion-pairing” effects, and, thus, ended up with an 
anomalous high concentration of Ca and Mg ions which are much higher than the model 
predicted value. Then he suggested that the identification of authigenic minerals in solid 
phases such as the lake’s sediments would provide information about the minerals 
controlling the brine evolution. 

Sundaram and Pareek (1995) reported that the quaternary landforms in the 
northern and eastern proximity of the Sambhar lake are products of depositional, 
structural and erosional processes. The depositional landforms are made up of fluvial and 


lacustrine and aeolian facies. The structural and erosional landforms are carved out of 
precambrian hard rocks. Of the two lacustrine facies the older is marked by the presence 
of saline deposits and the younger by saliferrous sediments. The depositional facies have 
evolved in a sequence which indicate slow variation in a predominantly arid climate. The 
presence of fluvial sediments forming the lake base contradicts the earlier observation 
which suggested an aeolian base for these lacustrine sediments. 

Apart fi’om the above geological studies, Gopal and Sharma (1994) have discussed 
about the biological diversity of the Sambhar lake. Large variety of phyto- and zoo- 
plankton, benthic invertebrates, fish and waterfowl existing in this wide range of salinity in 
the Sambhar lake. Particularly the seasonal behaviour of the blue-green algae is interesting 
and needs more attention so far as the brine evolution of the lake is concerned. 


A review of the existing work on the Samhar lake strongly suggests that there is a 
need to take up a regional study of the Sambhar lake area. The studies so far have 
concentrated either on the lake catchment or on the lake itself To understand the 
complexities of the physical and chemical evolution of the Sambhar lake, it was felt 
necessary to look at the geomorphological, sedimentological and sediment-mineralogical 
aspects simultaneously and to evaluate these data in the light of the available work on lake 
water chemistry and brine evolution. Keeping in view the above mentioned gap in the 
existing studies, the following objectives were envisaged for the present work; 


1. Interpretation of physical evolution of the Sambhar lake through geomorphological 
investigation using remote sensing techniques followed by field investigation. 

2. Interpretation of depositional environment and chemical evolution of the Sambhar lake 
through sedimentological and sediment-mineralogical studies. 


Both inaccessibility and the vast expanse of the lake render it impracticable to 
sample water and soil jfrom the entire lake bed. While the soft mud makes wading through 
the lake impossible, the shallow water impede boat movements. Thus, all the study 
conducted so far have, therefore, been biased towards sampling from few selective points 
mainly along the periphery of the lake. After understanding the pros and cons about the 
difficulties in accessing various parts of the lake, a plan was chalked out to orient the 
present work in a systematic way to achieve the objectives outlined above. The different 
work elements of the present study comprised of the following: 

(i) A thorough literature survey was carried out and all published maps and data 
related to Sambhar lake and the surroundings were evaluated. 

(ii) The first phase of field work was carried out in March, 1996 comprising of 
geomorphological investigations, collection of sediments and lake water samples and 
collection of relevant data from Sambhar Salt authorities. 


(iii) Next, the geomorphological investigations through digital image processing of 
satelhte data was taken up to understand the physical evolution of the Sambhar lake 
including the tectonic activities in the area. 

(iv) The sedimentological characteristics and mineral constituents of the sediment 
samples from different localities (see fig. 1.1) were studied in the laboratory. 

(v) The second phase of field work was carried out in October, 1996 for ground 
truth rectification of the remote sensing data analysis. Second set of sampling of sediments 
was also carried out followed by laboratory analysis. 

(vi) Finally, the field and laboratory results were synthesized to work out a 
physico-chemical model for the evolution of the Sambhar lake. 


The present thesis has been divided into five chapters. Chapter 1 includes a brief 
discussion on the saline lakes followed by a discussion on the available literature of the 
Sambhar lake, and the scope and objectives of the present study. Chapter 2 reviews the 
regional geology and geomorphology of the area. Chapter 3 includes the details of remote 
sensing techniques used for the interpretation of geomorphology and tectonic activities in 
the area in order to understand the physical evolution of the present day lake basin. 
Following this, the chapter 4 discusses in detail the procedure for sediment sampling and 
laboratory set up for sedimentological and clay mineralogy studies and the interpretation 
of the data obtained. In chapter 5 the entire work is summarised and major conclusions 
have been drawn. 





The Sambhar lake region falls on the mighty AravaUi mountain range which has 
received considerable attention of geoscientists since C. A. Racket first wrote a geological 
paper on it about hundred years ago. Studies made on diverse lines began to converge 
together, since then, permitting a regional synthesis. 

The AravaUi mountain range of Rajasthan and northern Gujarat comprises a 
number of fold belts of early and middle proterozoic age. These proterozoic fold belts, like 
other proterozoic fold belts of the world, evolved through the development of series of 
basins in which sediments and volcanics were laid down in several successive groups 
bounded by unconformities. Beginning around 250 miUion years back, the records of the 
proterozoic events in the AravaUi mountain range can be traced as late as c500 miUion 
years before present. 


The stratigraphic units forming this great mountain range include several 
imconformity-bound metasedimentary and metavolcanic units deposited successively over 
an ancient basement of plutonic rocks and para-gneisses. As a cover unit. Heron (1935, 

an ancient basement of plutonic rocks and para-gneisses. As a cover unit, Heron (1935, 
1953) recognised the “Aravalli system” as the oldest formation overlying the Banded 
Gneissic Complex (Roy, 1988), and the Bundelkhand gneiss (renamed by Pascoe, 1950, as 
the Berach granite) with a profound unconformity (Sen 1988). The “Raialo series” and the 
“Delhi system” are the two other cover sequences recognised by many workers. Table 2. 1 
shows the stratigraphic succession of the Precambrian rocks of the Aravalli mormtain 
range as recognised by Sen (1988). 

Table 2.1: Lithostratigraphic framework of the Aravalli Mountain Range 












(After Sen, 1988) 

A close scrutiny of the published maps of the Geological Survey of India (Fig. 2.1) 
reveals that the major parts of the Sambhar lake region is covered with quaternary 
deposits which, henceforth, conceal the facts about the exact structural and stratigraphic 
features. Nevertheless, it has been recognised that most of the outcrops west to this region 


Ill Granitic Rocks (Pre & Post/Late-Delhi Age) 
Delhi Supergroup 
Hill]] Aravalli Supergroup 

Banded Gneissic Complex & Berach Granite 
F---F Fault F-^ 

G.B.F. Great Boundary Fault ft ' 






,GRA . 














Fig, 2.1 Map showing the distribution of the Precambrian crystalline 
rocks in the Aravalli Mountain belt and its surrounding plains 
along with the study area Csimplifled after the published 
maps of the Geological Survey of India) 

belong to the Delhi supergroup, and the BGC and the Berach granite lies towards the 
south of the area. 

Zooming into the area of interest in the above discussed map (Fig. 2.1), it is clear 
that the Delhi supergroup, which constitutes the main edifice of the Aravalli range, 
occupies a narrow linear stretch in this part of the central Rajasthan. Sharma (1988) 
described the lithology of the Delhi supergroup rocks of this region as indicative of lower 
greenschist facies of regional metamorphism. The regional schistosity is considered 
synkinematic with folding on NE-SW axis and accompanied by low grade crystallisation of 
the metasediments (Basu, 1982; Gangopadhaya and Pyne, 1980; Singh, 1982). 

The Banded Gneissic Complex (BGC) exposed towards the south of the study area 
is considered as underlying the supracrustal of the Aravalli and Delhis, and is characterised 
by polymetamorphic mineral assemblages. The mineral assemblages comprising the rocks 
of the region suggests high pressure conditions during which the upper stability of 
muscovite had been reached (Sharma, 1988). Geothermometric and geobarometric 
estimations by Sharma (1988) suggest a pressure of 5.5-6 kb and temperature range of 
600°-700®C. The mineral assemblage comprising biotite-sillimanite- “melt”, in which melt 
is coarse grained quartz, plagioclase and microcline (Sharma, 1988). 

Three phases of folding have been recognised in the Delhi rocks (Mukhopadhaya 
and Dasgupta, 1978; Roday, 1979; Naha et al, 1984; Roy and Das, 1985). The DFl and 
DF2 folding episodes appear temporally very close to each other ^aneijee and Mitra, 
1977; Roy, 1988), both forming during the Delhi orogeny. Temporal reactions between 
the crystallisation and deformation suggest that the metamorphism of the Delhi 


supergroup rocks of this region was broadly coeval with the two phases of deformations 
(DFl and DF2), and the crystallisation outlasted the second deformation (Roy and Das, 

This part of the Delhi basin is characterised by several fossil grabens, horsts and 
grabens (Singh, 1988). The neotectonically active “Sambhar-Jaipur-Dausa wrench faulf’ 
has been presumed to be a paleotransform fault separating the “north Delhi basin”, which 
opened earlier from “south Delhi basin” opening at a later date (Roy, 1988). 


The Aravalli range, trending diagonally from north-east to south-west, opens out 
in a fan like fashion in Jaipur and Alwar districts. Described as the ridge and valley 
province, the Aravalli range in this part breached traversely to form a number of wind gaps 
including that near the Sambhar lake. An extensive pediplain truncates with the rocks 
belonging to the older sub divisions of the crust including the basement. The hill ranges 
are abruptly truncated both in the east and west, and show linear mountain fronts over 
long distances. The hill tops have preserved relict erosional surfaces at different levels with 
many of the hills showing first order topography, preserving anticlinal hills and synclinal 
valleys (Dassarma, 1988). The middle part of the pediplain is covered up with menacing 
aeolian sand shield and dunes of various morphological types. The drainage is essentially 
structurally controlled, showing preferred directions along NE-SW, E-W and NW-SE. A 
number of lakes including Sambhar lake occur in the eastern fringes of the wind gap and 
also across disorganised river courses. 


Remnants of an earlier planation surface occur as narrow linear stretches of 
different elevations on hill tops and rise in stepped terraces from the adjoining plains 
(Dassarma, 1988). The siufaces are longitudinally bounded by NE-SW and N-S faults and 
fractures, fortified with scarps and truncated by E-W cross faults. Effect of post planation 
uplift and faulting are manifested in the abundance of spectacular hanging valleys all of 
which show three distinct steps in their rise (Roy and Sen, 1983). 

In the present area effect of warping is discernible in the local segmented blocks 
bounded by longitudinal and cross faults (Sen and Sen, 1983). The anticlinal hills present 
in the area attains a maximum height at the middle and steadily decline in elevation both 
northward and southward by about 400 m till it is truncated by cross faults. A striking 
parallelism exists between the axes of warping and axes of DF3 folding in the area so 
much so that the maximum and the minimum elevation of the warped surface locally 
coincide with culminations and depressions of early fold axes where these are crossed by 
DF3 folding (Sen and Sen, 1983; Roy, 1983). Relevant in this context is the description of 
the DF3 formation as “broad warps” and “unaccompanied by internal deformation” 
(Roday, 1979), as “mild”, “extremely local” and “non penetrative” (Roy, 1983) and as 
“broad open cross-folds” which generated “wider spaced fractures” (Sen and Sen, 1983). 
In the present case, the prevalence of N-S fractures and scarps parallel to the compression 
axis, the general rejuvenation of NW-SE striking faults fully bears out a N-S directed 
compression axis (Ghosh and Viswanatham, 1991). Faults trending E-W and parallel to 
the axes of warping are possibly of tensional origin. 



The rejuvenated fault and fractures responsible for the segmentation of the 
erosional surface also control the overall drainage pattern of the area. The river Mendha 
flows straight along NE-SW stretches for major part of its course and shows N-S 
alignments particularly just before merging into the lake. All paleochannels of the Mendha 
river shows the initial dominance of NE-SW lineaments, and subsequent superposition of 
N-S through river capture. Differential uplift and tilting of N-S oriented blocks (Fig. 2.2) 
have been largely responsible for segmentation, ponding, disorganisation of rivers and 
formation of several saline lakes in the eastern fringe of the Aravalli range (Dassarma, 
1988). Most of the river channels display contrasting morphologies in adjacent segments, 
locally grafting relict meanderal loops of Roopangarh river to insurgent straight headway 
of river Mendha. As free meandering suggests absence of tectonic control, the cumulative 
evidences of meandering and recurrent river captures along successive preferred directions 
suggest periodic impulses (Williams and Clarke, 1995) 

Quaternary sedbnentation is mainly restricted to the marginal fault troughs fringing 
the uplifted western and eastern margin of the Aravalli range. The formation of the 
troughs and their rapid filling during the quaternary period indicate normal faulting 
accompanying block uplift; the drowning of the sediment suggests continuation of 
movements (Blatt, 1980). 

The deposition of linear alluvial valleys in successive parallel stretches suggests 
sedimentation in stepped graben in the Mendha valley (Fig. 2.2). The successive linear 
basins simulate stepped grabens formed by a series of antithetical faults heading towards 
the main fault (cf De Sitter, 1956). The linear and parallel vall^s in the Mendha basin 




Fig. 2.2 Neotectonic map of northeasforn Rajasthan’: prepared from landsat 
imageries followed by field checks. (After Dassarma, 1988) 

may indicate sedimentation either in Ridel shears formed oblique to the regional NE shear 
or in successive grabens formed by fault antithetic to the normal component of the original 
fault (Ghosh and Viswanatham, 1991). 


A summary of the structural, geomorphological, pedological and stratigraphic 
evidences in support of neotectonic movements in north-east Rajasthan has been furnished 
by Dassarma (1988). Sen and Sen (1983) suggested a post neogene uplift of the Aravalli 
range as a horst which, according to them, took place through re-activation of older 
lineaments, principally the Great boundary fault in the east and the Sardarsahar fault in the 
west. The structural evidences comprise truncated hill fronts with fault scarps, in alluvium 
and silcretes, and post orogenic bending, shattering and dragging of ridges with rotational 
effects to form transverse wind gaps. The pedological features include relict silcrete and 
calcrete skins on topographic highs, and anomalous depths of oxidation in contiguous 
blocks on opposite sides of the NE-SW trending faults (Dassarma, 1988). Swarms of 
parallel structurallycontrolled paleochannels, lineament-controlled river captures, river 
diversions, ponding and formation of salt lakes across disorganised channels and 
anomalous terracing in adjacent rivers are the important geomorphological features 
suggesting neotectonic activity in the area. The stratigraphic evidence of neotectonism 
includes differential accumulation of sediments in contiguous blocks with local “drowned” 
quaternary topography. 

Role of rotational movement was suggested for the formation of the conspicuous 
wind gaps, transverse to the Aravalli moimtain range between Sambhar lake and Klantli 


river where a series of NE-SW, NW-SE and N-S faults intersect each other (Dassarma, 
1985). A comparison of lineaments indicated by Mendha stepped grabens suggests a 
relative clockwise relation of the cmstal block east of the Aravalli orographic axis (Fig. 
2.2). The rotational process has also pulled apart a series of marginal depressions in the 
eastern fiinges of the gaps to form the Sambhar and some other smaller lakes with linear 

Anhert’s work (1970) on the relationship of denudation, relief and uplift suggests 
that if the uplift is only isostatic, then the main relief of any terrain will probably be 
reduced to 10 per cent of its original value after 30 million years. Thus, assuming the 
Aravalli range as an absolutely stable region at least since Mesozoic, the present mean 
relief of 300m would imply a relief of 3,000 m in Oligocene and 30,000 m in Paleocene. 
Such high relief would suggest that the present topographic and geomorphic features of 
the Aravalli range to be interpreted as rejuvenated feature of late Tertiary or early 
Quaternary age. 

A possible driving force for the post-orogeny movements could be during the late 
Pliocene early Pleistocene time when the continental cmsts of the Indian and Eurasian 
plates became locked (Sen and Sen, 1983). The differential movement caused rotation of 
some of the blocks along reactivated faults, ripping open in the process linear depression 
to form Quaternary sedimentation basins and lakes. The absence of sediments earlier than 
Quaternary in all these depressions provides stratigraphic evidence for recent movements 
(Roy and Sen, 1983). The movements were perhaps impulsive as suggested by anomalous 
terracing and ephemeral meandering phases of paleochannels. 





Remote sensing by definition is the method of acquiring data about an object 
without any physical contact with the object itself It has two facets; the technology of 
obtaining data through a device which is located at a distance Jfrom the object, and the 
analysis of the data for interpreting the physical attributes of the object, both these aspects 
being intimately inter-linked with each other (Gupta, 1991). 

The fundamental principle of remote sensing is that, depending upon the physical 
and/or compositional attributes of the object certain intensity of light reflects within a 
particular range of electromagnetic spectrum (Fig. 3.1). Thus using information firom one 
or more wavelength ranges it may be possible to differentiate between deferent type of 
objects and map their distribution on the ground. 

Geomorphology deals with the study of landforms and landscapes, including their 
description, t 5 q)e and genesis. Landform is the end product resulting fi'om the interaction 
of the natural surface agencies and the type of rock (Bloom, 1969). It depends on three 
main factors such as (a) climatic setting, including its variation in the past (b) lithology and 
structure, and (c) the time span involved. 


Fig. 3.1 T\-pical spectral reflectance curves for 
selected common natural objects — water, 
vegetation, soil and limonite 

(source Gupta, 1991) 

One of the widest application of remote sensing data has been in the field of 
geomorphology, because remotely sensed data products give direct information about the 
surface feature on the earth. Also landform features can be well studied in a regional scale 
using synoptic coverage provided by remotely sensed satellite data, rather than in the field. 


The digital remote sensing data of IRS-IB LISS2 used in the present work is 
provided by NRSA, Hyderabad. Linear Image Self Scanning (LISS) pay load of Indian 
remote sensing satellite IRS- IB consists of three solid state cameras: low resolution 
(72.5m) LISSl, and medium resolution (36.25m) LISS2A and LISS2B. LISSl provides a 
swath of 148 km, while the composite swath of LISS2A and LISS2B is 145 km. IRS 
satellites are placed in a 904 Km polar sun-S3mchronous orbit with an orbit period of 103 
minutes. The satellites return to their original orbit trace every 22 days enabling repeat 
collection of data over the same place and at the same local time. The detail specifications 
of the data used are listed in Table 3.1. 

Table 3.1: Specification of the data used 

Production identification code 


Date of product generation 








Sub-scene/SOI map sheet No. 


Date of pass 

21 Nov. 1995 

Input scene centre latitude (degrees) 


Input scene centre longitude (degrees) 


Inter pixel distance (meters) 


Inter line distance (meters) 


Band Interleaving Indicator 


Band Nos. 


Record level of volume directory file 

360 bytes 

Record level of leader file 

3960 bytes 

Record level of imagery file 

2520 l^es 


LISS cameras observe radiance reflected fi'om earth’s surface in four bands of the 
following wavelength; 0.45 |am to 0.52|j,m (Band 1), 0.52 jim to 0.59 gm (Band 2), 0.62 
|a,m to 0.68 |jLm (Band 3) and 0.77 jim to 0.86 |xm (Band 4). Table 3.2 lists the major 
characteristics of IRS bands and their applications. 

Table 3.2: IRS Spectral Bands and their principal applications 

Band No. 

Spectral range(um) 

Spectral location 

Principal applications 




Sensitive to sedimentation, 

deciduous/coniferous forest cover 





Green reflectance of healthy vegetation 




Sensitive to chlorophyll absorption by 
vegetation, dhOferentiation of soil and 
geological boundary 




Sensitive to green biomes and moisture in 
vegetation, land water contrast 

A micro processor based digitally controlled auto tracking fi'equency colour 
monitor (HC 3925 series) firom Mtsubishi Electric Corporation was used to display the 
images from a MS-DOS with 486 DX main processor and 256 Kb kache memory. The 
main software used in this study was ERDAS and sometimes the help of another PC based 
software namely IDRIS! was taken. ERDAS (Earth Resource Data Analysis System) is an 
efficient PC based GIS package developed by ERDAS Inc. Ltd. This package is widely 
used by various agencies round the globe for analysis of satellite data for geology, 
forestry, pedology, agriculture, environmental sciences, etc. This soflware contains 
various programs and algorithms to enhance the available satellite data. The multi-band 


digital data of the Sambhar lake area were processed using standard image processing 
techmques. Various image transformations were applied to enhance the geomorphological 
features present in the study area which enabled to interpret the physical evolution of the 
Sambhar lake and its surrounding. The results obtained through image processing were 
verified through actual field visits. The details of the image processing techniques and their 
results are presented next. 


Digital image processing involves the manipulation and interpretation of digital 
images with the aid of a computer. The digital image is fed into a computer one pixel at a 
time. The computer is programmed to insert these data into an equation, or series of 
equations, and then store the result of the computation for each pixel. These results form a 
new digital image that may be displayed or recorded in pictorial format or may itself be 
further manipulated by additional programs. The possible forms of digital image 
manipulation are literally infinite (Lillesand and Kiefer, 1994). Image enhancement, one 
category of digital image processing, is the process of making an image more 
interpretable for a particular application. Enhancement can make important features of, 
raw, remotely sensed data more interpretable to the human eye. Enhancement techniques 
are often used to study and locate areas and objects on the ground and deriving useful 
information from images. The various enhancement techniques used in this present work 
include spectral enhancement, spatial enhancement, multi-band enhancement and principal 
component transformation. 


3.3.1. Spectral Enhancement 

Spectral enhancement deals with the individual values of the pixel in the image. 
The goal of spectral enhancement is to make certain features more visible in an imagery by 
bringing out the contrast. Depending upon the features to be extracted and the band in 
which they appear, spectral enhancements such as linear contrast stretching was applied to 
a particular band data. 

Single band images were extracted from multiband data and displayed on 
monochrome plmn after contrast stretching. Each band exaggerates certain features while 
suppressing others depending on the reflectance value of the surface cover. Band 3 being 
sensitive to chlorophyll highlights the vegetation as dark gray patches (Plate 3.1). It helps 
in differentiating soil types and delineating geological boundaries. In this band fresh and 
salt encrusted sand is represented by the brightest tone while old sand is represented by 
lighter shades of gray. Contrast stretched image of Band 4 (Plate 3.2) is elaborating the 
land water contrast, and green biomass and moisture being sensitive to this wave band 
appearing as distinct patches of lighter tone in the image. The manmade structures are 
delineated from their distinct pattern and contrasting tone with the surrounding in the band 
3 (Plate 3.1) image. In band 1 image the fresh alluvial sediments are appearing as brightest 
tone while the aeolian and the old alluvial sediments are showing lighter tone. 

3.3.2. Spatial Enhancement 

Spatial enhancement technique emphasize or de-emphasize image data of various 
spatial frequencies, which is the difference between the highest and lowest values of a 


contiguous set of pixels. Jensen (1986) defines it as “ the number of changes in brightness 
values per unit distance for any particular part of an image”. The process of aver aging 
small sets of pixels across an image is called convolution filtering which is used to change 
the spatial fi-equency characteristics of an image (Jensen, 1986). For example, a Zero 
spatial firequency produces a flat image where every pixel has the same value; a Low 
spatial fi’equency is an image consisting of smoothly varying gray scale, and high spatial 
frequency results an image consisting of a checkerboard of black and white pixels. 

The commonly used convolution kernel are high frequency kernel, zero-sum kernel 
and low frequency kernel, as given bellow: 

-1 -1 -1 

-1 -1 -1 

1 1 1 

-1 16 -1 

1 -2 1 

1 1 1 

-1 -1 -1 

1 1 1 

1 1 1 

High fi-equency kernel 

Zero-sum kernel 

Low fi-equency kernel 

In the present work, high firequency kernel and zero sum kernel were used. 
High firequency kernel serves as an edge enhancer, since they bring out the edges between 
homogeneous groups of pixels. Unlike edge detectors (such as zero sum kernels), they 
only highlight edges, they do not necessarily eliminate other features. While low fi-equency 
kernel simply averages the value of the pixels, causing them to be more homogeneous 
(homogeneity in low spatial fi-equency). The resulting image looks more smooth or more 
blurred. The filtered images of bands 1, 2 and 4 were loaded separately in planes 1, 2 and 
3 respectively to make the colour composite. 


The edge enhancing high frequency filter produced an image (Plate 3.3) which 
illustrates the geological and structural features as well as topography. The quartzite 
ridges of this area are very well identified as narrow linear patches of white colour which 
are accompanied by dark gray shades indicating foothills. The relief differences of different 
topographic features are indicated by the shades of white, brown and pink. The zero sxun 
edge, on the other hand, produced an image which helps in demarcating the edges 

3.3.3. Multi-Band Enhancement 

A peculiar aspect of remote sensing is that it provides data in multiple spectral 
bands which can be super-imposed over one another to deduce information not readily 
seen on a single image (Gupta, 1991). Radiometrically corrected IRS LISS2 data of bands 
1, 2 and 4 were loaded at three different planes (1,2 and 3). Red was associated with band 
4, green with band 2 and blue with band 1 to generate a False Colour Composite (FCC). 
The histogram obtained for each band image was stretched separately and superimposed 
(1:1 ratio). 

The FCC (Plate 3.4) provides a regional view with respect to geology, 
topography, drainage and vegetation in different shades of colours. The shades of dark 
brown represent greatest height corresponding to quartzite and quartzoschist ridges and 
light brown colour represents sand dunes and sand mounds of different heights. Most of 
the drainage seen on the image are dry channels and are depicted by the shades of white 
and while the plaeo-channels having high moisture content and scanty vegetation 


Plate 3.4 Multiband enhancement — standard FCC 
prepared by coding band 4,2 and 1 data on red, 
green and blue colour reqjectively 

rte 3.3 Spatially enhanced image of band 4,2,1 
ng edge enhancing high frequency kernel 

developed on it are depicted by ink blue shades. Apart from the main Sambhar lakes, a few 
smaller lakes are also seen which seem to be connected by one of the old channels. Thickly 
vegetated area are represented by different shades of red colour. The different part of lake 
are well depicted in this FCC as main lake body shows deep blue colour while the 
reservoir and the bittern part of the lake are showing green and black colours respectively. 

The FCC shown in Plate 3.5 was obtained by coding band 4 with red, band 3 with 
green band 1 with blue shows a better contrast between different land covers. The Aravalli 
ridges are represented by narrow elongated zones of deep brown colour. While the alluvial 
fans developed on the footslopes are represented by lighter shades of brown (Plate 3.5). 
Drainage and salt encrusted sand are represented by brightest tone. Dark green colour 
represent vegetation while the lighter shades of green are representing aeolian sand and 
sand dunes of different altitude. Different parts of the lake viz. the main lake body, the 
reservoir and the bittern dumped eastern part are also well demarcated in the imagery 
(Plate 3.5). 

Principal Component Transformation (PCT) is a technique designed to remove or 
reduce the redundancy in multispectral data. This transformation may be applied either as 
an enhancement operation prior to visual interpretation of the data or as a pre processing 
procedure prior to the automated classification of the data (LiUesand and Kiefer, 1994). If 
employed in the later context , the transformation generally increases the computational 
efficiency of the classification process because PCT may result in a reduction in the 
dimensionality of the original data set (Gupta, 1991). 


Using PCX technique a new set of coordinate axes were fitted to the image data. 
The first new axis or component represents an orientation which shows mayimnm variance 
accounted for that axis. Subsequent components (axes) account for successively smaller 
portions of the remaining variance. The covariance matrix of the PCX is given in Table 

An enhanced colour composite (Plate 3.6) was produced by loading the images 
(PCI, PC2 and PC3) in planes 1, 2 and 3 respectively. The product of PCX (Plate 3.6) 
highlights the characteristic differences between fluvial and aeolian sedimentary units, and 
drainage of past and present origin. The Aravalli hills are represented by narrow linear 
patches of dark green colour and Lighter shades of green represents the Quaternary fluvial 
deposits of various generations while the aeolian deposits of various generations is 
represent by light pink (old) and white (younger). Past drainage and filled up depressions 
are represented by narrow linear patches of light greenish yellow colour. Present drainage 
is prominently represented by the narrow linear patches of blue colour. Salt encrusted 
alluvial sands spreading over the deltas and dried channels are represented by different 
shades of pink. 

Table 3.3 : Covariance Matrix of the Principal Components 



























Plate 3.5 Multiband enhancement— FCC Plate 3.6 Principal Component Transformation 

prepared by coding band 4,3 and 1 data on re4 (P^T) — enhanced colour composite by loading the 

green and blue colour respectively images (PCI, PC2 and PCS) on planes red, green 

— and blue respectively 


Plate 3.7 Band ratioing— band ratioed image using Plate 3.8 Band ratioing— band ratioed image using 


Z4+X3 X1+X2 

3.3.4. Band Ratioing 

Band ratioing is an extremely useful procedure for enhancing features on the multi 
spectral images. It is used to reduce the variable effects of illumination condition and 
topography (Gupta, 1991). The new digital image is constructed by computing ratio of 
DN values in two or more input images, pixel by pixel. The general concept can be 
formulated as; 


DNnew = m\ — — — — +n Where, 


DNa= DN values in input image A 
DNb= DN values in input image B 

Ki and K 2 = factors which take care of path radiance present in two input 

images, and 

m and n= scaling factors for gray range 

Image for complex ratio parameters including addition, subtraction, multiplication 
and double ratios etc. can be generated similarly. The resulting ratio image can again be 
contrast stretched or used as a component for other enhancements. 

Ratioing is done mainly taking only band 4 and band 3 data as in band 4 vegetation 
has a lower rachance than soil and in band 3 the opposite is true. The mathematics used to 
delineate natural stream flow and vegetation patterns which align with geological 
X4 — X'i 

structures is given as where X represents the DN values of that band. 


This ratioed image after contrast stretching and enhancement (Plate 3.7) illustrates 
thickly vegetated area with white patches while the streams are represented by lin ear 
patches of gray shades. 

To suppress the minute details of the area and to exaggerate the land, water and 
vegetation in a broad scale another mathematics was also used taking the band 1 and 2 

data, which is given as 

X\ + XI' 

The resultant image (Plate 3.8) shows water and salt encrusted sand as brightest 
tone while the vegetation are represented with the dark patches. The aeolian sand plain is 
showing an intermediate tone. In this image the present day channel is. well demarcated 
which helped in determining the drainage pattern and thus in preparing tectonic map. 

3.3.5 Classification 

Multispectral classification is the process of sorting pixels into a finite number of 
individual classes, categories of data, based on their data file values. If a pbcel satisfies a 
certain set of criteria, the pixel is assigned to the class that correspond to that criteria. 

Depending upon the type of information to be extracted from original data, classes 
may be associated with known features on the ground or may simply represent area that 
“look different” to the computer. An example of a classified image is a land cover map, 
showing vegetation, bare land, pasture, urban areas, etc. 

Pattern recognition is the science-and-art of finding meaningfiil patterns in data 
which can be extracted through classification. By spatially and spectrally enhancing an 
image, pattern recognition can be performed visually or through a computer system. The 
later case is more scientific. 

The study area being a typical semi-arid region poses a serious problem in 
classifying the satellite data through known image processing techniques. The 
measurement of vegetation cover in this semi-arid region is complicated by variability in 
the soil reflectance as well as spectral interactions between the sparse plant canopies and 
the soil. This problem is well elaborated and discussed by Ray and Murray (1996) for 
desert vegetation. To minimise the problem reasonably two separate classification 
algorithms were used to enhance the classes in the lake and those in land (mainly soil) 
separately. One is the clustering algorithm and the other is the ISODATA algorithm. 
Though in both the cases the mapping of the vegetation as different class was not possible 
to its best extent, but it has solved the problem of enhancing the classes both in land and in 
lake separately. 

In clustering algorithm, statistics are derived fi'om the spectral characteristic of all 
pixels in an image and then, the pixels are sorted based on mathematical criteria. The 
statistical method of unsupervised training takes into accoimt the homogeneity of 
neighbouring group of pixels, instead of considering individual pixels equally (Lillesand 
and Kiefer, 1994). The algorithm only uses 3x3 sets of contiguous pkels that have similar 
measurement vectors. Any other 3x3 windows of pixels are discarded. The assumption 

behind the used algorithm is that contiguous, homogeneous pixels usually indicate a spatial 
pattern within the data (such as land cover type) that is worth classifying. 

The output clustered image (Plate 3.9) represent the descrete hills of Aravalli 
range as dark green patches surroimded by alluvial fans showing blue patches. S till light er 
blue patches represent sand dunes and mounds of different height. The pink patches which 
spread all over either of the river channels and their deltas along the lake shore represent 
moist salt encrusted sand having higher reflectivity. The red tints within this pink patches 
represent relatively higher altitude. The aeolian sand is represented with patches of lighter 
pink (old sand) and yellow patches (new sand and sand dunes). The pink patches outside 
either of the river channel and lake shore represent paleo-channels. The three major parts 
of the lake body is distinctly separated with brine reservoir being deep blue in colour while 
bittern is represented with the darkest tone. The main lake body shows intermediate shade 
of blue. The clusters of green patches all over the area except those representing linear 
trend of the Aravalli hills are representing vegetation. 

ISODATA stands for “ interactive self-organised data analysis technique”. It is a 
“iterative” classification technique and in that way it repeatedly performs an entire 
classification and recalculate statistics. “ Self organising” refers to the way in which it 
locates clusters with minimum user input (Lillesand and Kiefer, 1994). The ISODATA 
method is somewhat similar to the sequential method discussed above. Th^ both use 
minimum spectral distance to assign a cluster for each candidate pixel. The major 
differences are that the ISODATA process begins with a specified number of arbitrary 
cluster mftans and then it process repetitively, so that those arbitrary means will shift to 


Plate 3.9 Classified image using statistical 
clustering of pixels of PCs 

Plate 3.10 ISODATA classified image showing 
various classes in lake water 

the means of the cluster in the data. Because the ISODATA method is iterative, it is not 
biased to the top of the data file, as are the one-pass clustering algorithms. 

A image file is created, which gives results similar to using a miriimnm distance 
classifier on the signatures that are created. The image file (Plate 3.10) shows a number of 
classes inside the lake water. These classes probably representing different s alinit y level 
and/or difference in sediment dispersion pattern, as the gradient of the lake bottom does 
not suggest that these classes are because of the depth of the water column. The pink 
patches in the image (Plate 3.10) represents salt encrusted sands. The light gray patch 
rimming the lake is indicating mud. The two different classes in the lake water needs 
greater attention. However, this can be considered as due to cumulative effect of different 
physical and chemical conditions in different parts of the lake. 


Multispectral remote sensing data has shown tremendous potential for application 
in various branches of geology- geomorphology, structural geology, lithology, ground 
water and geo-environmental studies, etc. The interpretation of image data, whether 
enhanced or simple products, is carried out using elements of photorecognition- tone or 
colour, texture, shape, size and pattern; and geotechmcal elements- landform, drainage, 
vegetation, lanH use and soil. The rock attributes (structure, lithology, etc.) and physical 
processes (climatic setting, weathering and erosional agencies) operating in a region over 
a time, govern the nature and appearance of landscape- relief, topography, vegetation. 

drainage, soil, etc. (Fig. 3.2). 






• ON 
' Images 

Fig. 3.2 Conceptual diagram showing various 
parameters with their sequence in nature 



The study area is an classic example of manifestation of aeolian, fluvial and 
lacustrine processes and landforms. Of these wind action is certainly predominant while 
surface water processes assume importance during the monsoon period. The effect of 
lacustrine process is marginal as the lake is shallow in nature; maximum depth being 
slightly more than Im during monsoon, and is practically devoid of any current and tidal 
action. Further the area shows numerous evidences of neotectonism. Using the digitally 
processed images, toposheets and followed by ground checks, thematic maps were 
prepared with respect to geomorphology, structure, lithology, etc. of the Sambhar lake 
area and its surroundings (Figs. 3.3 and 3.4). The geomorphological processes and the 
landforms related to these three different enviromnents are discussed separately. 

3.4.1 Fluvial Processes and Landforms 

The fluvial action in the Sambhar lake catchment is performed by two major rivers 
of this locality namely, Mendha and Rupangarh, flowing fi'om NNE and SSW direction 
respectively. As the flow of water in these two channel systems are not maintained 
throughout the year, a better terminology for these two inlet systems to the lake would be 
“Wadi”. Many other small streams are debauching from the Aravalli hills and continuing 
their flow till their course is chocked up by their own load or by the migrating wind 
derived sand bodies or dimes. The entire drainage systems of this area is structurally 
controlled as suggested by their drainage patterns. 

Amongst the two major wadis, the channel of the Mendha is wider than that of the 
Rupangarh. Again the channel of the Mendha is narrowing down towards the lake. The 


r»g. 3.3 Geomorpholo^cal map prepared from digitally 

processed satellite images aided by Toposheets and 
around cher.k-R 

Mendha is approaching the lake in a straight course from NE direction till its course is 
forced to follow a N-S alignment by the presence of a fault. The Mendha is maintaining a 
straight course throughout the area suggesting a strong structural control. Near the bridge 
(Plate 3.7 and 3.8) the channel of the Mendha is found to be chocked up by sand dunes of 
barchan type and the satellite image shows that the Mendha channel disappears into the 
ground here. But, the presence of deltaic deposits in northern shore of the lake suggest 
that immediately after rain the Mendha probably gets reactivated and then starts to flow 
following the rain. The damming of the water in its channel would cause the accumulation 
of water and then a flash flood removing all aeolian sands deposited along its course. The 
presence of flood plain deposits on the western bank of the river (Fig. 3.3) supports this 

In contrast to the straight course of the Mendha, the Rupangarh has a cumulative 
expression of meandering and river capture prior to its final merge into the lake. This kind 
of drainage pattern suggests the neotectonic activity in periodic pulses (William and 
Clarke, 1995). 

Both the rivers have a number of paleochaimels suggesting lateral shifting of 
channels in the recent past (see Fig. 3.3). The general shifting pattern of the channels 
present towards the North of the lake is rotational, being rotated in the anticlockwise 
direction. But the channel of the Rupangarh which is present towards the south of the lake 
shows a general shifting towards NW direction. The differential channel migration pattern 
on either side of the lake suggests the presence of two separate tectonic blocks on either 
side of the Sambhar lake which are tilting and/or moving individually. This may be the 


cause of the fonnation of the Sambhar lake basin. Apart from the above discussed 
paleochannels there are some paleochannels present in the southern and eastern side of the 
lake which do not seem to be related to either of the two m^or drainage systems of the 
area. These are showing almost E-W trend, and the one present parallel to the southern 
edge of the Sambhar lake has two small lakes within it’s channel (Plate 3.4 and 3.5; Fig. 

Apart from the drainage systems of the area, the other fluvial landform present in 
this area are the coalescing alluvial fans (Plate 3.11). These are formed by the streams 
debauching from the Aravalli range from all the sides. The streams debauching from the 
Aravalli peaks nearer to the lake continue their flow into the lake thus feeding it with 
water and sediments directly from the hills. The coalescing alluvial fans are covering a 
huge area in the foothills. The small stream course are of swinging pattern on these alluvial 
fans as noticed in this satellite imagery. 

3.4.2 Aeolian Processes and Landforms 

Whatsoever their origin, the sand grains, in this region are transported over the 
surface by the wind, until the surface wind velocity is reduced sufBciently for them to 
come to rest. On a large scale, this is happening in the continental Sambhar basin and in 
other depressions present in the area and in the topographic traps formed by the Aravalli 
range (Plate 3.12). On a small scale, sand grains come to rest in the comparative 
protection of the lee side of the boulders and/or vegetation, or in the shelter of a river 
bank as seen in the field. 


Plate 3.12 Ground photograph showing alluvial 
&ns and aeolian sand deposits near lake bed, 

As considerable amount of sands are available, sand dunes are developed at many 
locations with deposition taking place mostly on the lee slopes of crescent or barchan 
dunes (Plate 3.13). As loose sands are readily transported, most major accumulations of 
sand dunes- the sand seas of this region- are occurring in depressions, as seen on the 
Sambhar basin and in the adjoining lowland areas surrounded by hills on both the sides 
(Plate 3.12). 

Barchans sand dunes are seen on the Mendha river bed which suggests an interplay 
of aeolian and fluvial processes; wind is also rearran^g the available alluvial sands on the 
river bed in dune forms (Plate 3.13). So far as the wind erosion is concerned, many 
faceted and polished blocks, and pedestal blocks were observed all along the Aravalli 
range during the field tour. The wind abrasion and attrition is producing large amount of 
sand firom the nearby rock exposures. 

3.4.3 Lacustrine Processes and Landforms 

The Sambhar lake present in this region is the most important and extensive of the 
saline lakes of Rajasthan. It occupies a huge area of about 233 km^ with its long axis 
aligned approximately in the East-West direction. The lake fills to a depth of over Im 
during the rainy season but dries out completely during the summer months. 

The lakft interior is inaccessible due to the presence of swamps developed in the 
peripheral region, as observed during the field tour. The outermost part of the swamps 
dried up during the summer months and can be accessed. As the depth of the lake is not 
conducive for any VinH of mode of transport in the water, the only possibility of collecting 


samples is from the periphery of the lake, thus it is very difScult to study various aspects 
at the middle part of the lake. 

The eastern part of the lake forms a reservoir for salt production and is detached 
from the main body through a dam connecting Jhapok and Gudha sites (Plate 3.14). Most 
researchers have collected lake water samples near the gate of the dam due to easy 
accessibility. A part of the reservoir is divided into no of kyars where the lake water is 
evaporated to achieve the required density level for salt production. The water samples 
from these kyars give a anomalously high concentration of salt. Also, a large population of 
algae are reported in the kyar. The biogeochemical activity in the kyar and brine reservoir 
are well reflected in the FCCs with a different colour expression. The mud excavated from 
the floor of the kyar i,e. “bittern” are dumped further eastern part of the lake and is well 
demarcated in the FCCs (Plate 3.4 and 3.5). The dried surface of the lake in the months of 
the summer shows polygonal cracking with salt mixed with clay filling the cracks. 

A terrace is reported at the western periphery of the lake. Due to the shallow 
depths the tidal and current action are not manifested in the lake, however seasonal 
variation of water column is observed, as reported by the local people. Lake is mainly fed 
by Mendha, Rupangarh and other small streams. But, the ground water contribution to the 
lake can not be ruled out as there are number of “losing” streams present in the area., 
forming potential ground water recharge zones. Further, presence of fractures, faults and 
other lineaments would also encourage grormd water recharge to enter into the lake. The 
lake deposits consist of grain size distribution of sands and clays earned by the rivers to 
the lake and those transported by the wind. The lake sediments suggest that they are 


Plate 3.13 Ground photograph showing barchan 
type sand dunes developed on Mendha river 


Plate 3.14 Standard FCC showing different 
parts of the Sambhar lake 




derived froin various sources and also transported with various modes (discussed later in 
chapter 4). The in-situ clay mineral formation in the lake sediment is inferred and an 
attempt has been made to elaborate the role of the organism in this clay formation 
(discussed later in chapter 4). 

Gigantic sand mounds and dunes are reported at the SE fringe of the lake. These 
dunes seem to have been stabilised due to kankar formation at their bases (pedogenic?) 
and growth of vegetation. It has been inferred these sand were once the part of the lake 
sediment and in the later period excavated and transported through wind deflation. The 
high wind through the gap in the Aravalli from the west of the lake and the presence of the 
above discussed sand dunes and mounds on the eastern periphery of the lake strongly 
support this theory. 

3.4.4 Geomorphology and Tectonics 

The satellite imageries and field observations reveal a number of tectonic features 
present in the area which are manifested in the overall geomorphic expression of the area. 
The typical drainage pattern, off-setting of rivers, linear alignment of small lakes and 
disrupted Aravalli range are only some of the features strongly suggestive of tectonic 
activity in the area. A structural (tectonic) map (Fig. 3.4) has been prepared by visual 
interpretation of the band 4 image (Plate 3.2) and the directional filter product of band 3 
and band 4 (Plate 3.3). Various structural units have been high lighted and the most 
possible directions of movements is shown in the map (Fig. 3.4). The visual interpretation 
and field observations suggest that the whole area can be divided into four tectonic blocks. 


Fig. 3.4 Structural map of the study area prepared from 
filtered (High Pass Kernel) satellite image of the 
Sambhar area 

One is present towards the west of the mountain range while other three can be 
demarcated by considering that the E-W elongated Sambhar lake is on the block which is 
sandwiched between two separate blocks towards its North and South respectively. These 
three blocks may have been rotating, moving and/or tilting individually. There is a growing 
debate about the origin of the Sambhar basin as a result of such tectonic activities. From 
the shifting of the channel of Mendha river it has been inferred that the block towards the 
north of the lake is tilting in the NW direction, while the block towards the south is also 
unstable as indicated by the anomalous drmnage pattern of the Rupangarh river. The 
independent movement of these two blocks might have led to the formation of some weak 
linear zone which eventually developed into Sambhar lake basin. The wind blovring from 
the gap in the Aravalli has done a great job through deflation of the lose material produced 
by tectonic activity along this linear zone, finally making a great depression in the sand. 
Apart from the independent movement of the individual blocks, a number of faults have 
been reported (Fig. 3.4) in each block. The drainage pattern in the blocks on either side of 
the lake manifests a number of faults. The broken hills of the Aravalli range also indicate 
‘faulting activity. The upliflment of the Aravalli along normal faults (Plate 3.15) at places 
is also noticed in the field. There are more than one generation fault reported in each block 
which is finally producing a superimposed pattern. 


The chapter has emphasized the potential of remote sensing techniques for 
landform map pin g and interpretation of various geomorphic processes operating in the 
region. On regional scale, the aeolian process are indeed dominating as reflected by 


Plate 3.15 Ground photograph showing a part of an 
entrenched stream with its thick soil embankment 
suggesting upUftment of Aravalli hill at the 

widespread sand dunees, scanty vegitation and other related landfomis. The fluvial action 
is limited spatially as well as temporally. Most commonly small scale alluvial fans have 
developed on the foot slope of the Aravally hills and a number of dry channels can be seen 
in the Sambhar lake catchment which are activated only during the moonsoon period that 
too for a very small time span. The lacustrine processes and their products are governed 
extensively by aeolian and fluvial processes and their products are governed extensively by 
aeolian and fluvial process in the region, and of course the climatic conditions. Due to very 
shallow, but areally extesive, nature of lake, typical lacustrine landfomis such as shoreline 
features, terraces, etc. have not developed. More appropriately, the sambhar lake can be 
classified as a ‘"wetland” a thin sheet of water is surrounded by swamps. The easternmost 
rocks and sediments of the lake form the focus of the present study and their analysis is 
treated in next chapter. 





A number of studies have been carried out till date to describe the physicochemical 
evolution of the Sambhar lake. Various workers have adopted different techniques to 
explore the possible source of the salt in Sambhar lake. Geochemical and isotopic study 
conducted by Yadav (1995), radiocarbon determinations of the lacustrine deposits of 
Sambhar by Singh et a/.(1972), and the geochemical study by Godebole (1952) are some 
of them. The detailed geomorphological study of the Sambhar lake and the surroundings 
(see chapter 3) reveals that the source of the Sambhar lake sediments are the exposed 
Aravalli rocks and the thick aeolian and alluvial plain surrounding the lake. The major 
geomorphologic agents transporting the sediments to the lake bed are the rivers flowing 
into the lake and the wind blowing through the gap in the Aravalli. Prior to their rest on 
the lake bed, the sediments undergo a series of physical and chemical weathering action 
which are a function of climate, types of weathering agents and neotectonic activity of the 

Keeping in view the conflicting theories regarding the evolution of Sambhar lake 
brine, the present study focuses on mineralogical study of the source rock and fluvio- 
lacustrine sediments of the area to understand the sediment water interaction phenomenon. 


Keeping the objectives in mind, the rock and sediment samples were collected 
from various locations (see Fig. 1.1) in the study area. The fresh rocks as well as the 
weathered part of it were detached from the exposed Aravalli system and then hammered 
into small rectangular shape for safe carrying and proper utilisation in the laboratory. The 
lake and river sediments were sampled from different depths through angering at each 
location (see Fig. 1.1), and were immediately packed in polythene bags to avoid 
contamination. Then, the samples were labeled on the spot for proper identification in the 
laboratory. The analysis of source rock samples mainly involved microscopic identification 
of rock type and the alteration products of the constituent minerals. The results obtained 
were confirmed through XRD analysis of the powdered rock samples. The sediments were 
subjected to grmn size analysis through dry sieving and clay mineralogical analysis through 

4.3,1 Methodology 

Both the megascopic and microscopic identification of rock were carried out. The 
thin section of the rock samples were prepared and studied under LEITZ, LABORLUX II 
POLS microscope under different magnifications. The optical properties like colour, 
shape, cleavage pattern, extinction angle, twinning and interference colours were studied 
in order to identify the mineral constituents of the rock and their alteration product. 


* I- T,. KANPUR 

4.3.2 Results and interpretation 

The physical observation of the rock samples revealed a variety of features 
manifesting weathering phenomenon. The schistose rock exposed on the lake bed is 
severely affected by salt weathering so much so that near Jhapok, fragments of orthoclase 
and mica are coming off from the rock mass and ready for further action, and pyrite are 
deposited in the form of small yellow colour tints on the surface of the rock. The red, 
brown and yellow hues to the surface of all rock samples from nearby hills suggest 
retention of iron in the ferric state. This impressions on the rock surface needs further 
explanation, and for which the microscopic study was carried out. 

Thin sections of the rocks from the R1 and R2 exposures (see Fig 1.1) shows that 
these are chiefly composed of quartz, feldspars and mica in different proportions. Samples 
from R1 has low mica content while the rock from R2 has abundant mica forming layers in 
between quartzo-feldspathic layer (Plate 4.1a and 4.1b) suggesting the rock as 
quartzoschist. Sericitisation along feldspar-mica interface is observed under microscope 
(Plate 4.2). Feldspars are mainly orthoclases of microcline and low albite type as 
suggested by their cross hatch and albite twinning respectively under microscope (Plate 
4.3). The XRD analysis of these rock samples reveals that the alteration product is largely 
illite with mica-montmorillonite mixed layered clays. 

The alteration phenomenon and the characteristic alteration product i,e. illite 
present in the rock samples from the exposed Aravalli hills suggest that topography and 


Plate 4.1 (a and b) Photomicrograph showing the pressence of mica in the 
rock(450X, crossed nicol) 

Plate 4.2 (a andb) Sericitisation along the feldspar edge 
(450X, crossed nicol) 

iate 4.3 Plagioclase feldspar (Albite) showing twinning 
(450X, c rosse d nicol) 

Plate 4.4 Doleritic rock showing plagioclase laths surrounded by 
chloritised augite(450X, crossed nicol) 

climate are the key factors controlling the rate of chemical weathering and the nature of 
weathering products. Even though the Aravallis have steep slopes, the fractures and 
ruptures in the rock mass cause suflSdent infiltration of rain water. However, as the 
climate of this area is typically semi-arid and the evaporation exceeds precipitation to 
certain extent, the water penetrating the rock returns to the surface during the ensuing 
long dry-spell and ultimately and is ultimately evaporated. As a result, the soluble 
constituents i,e Na'*’, K"^, etc. are not removed to their best extent and thus reactions are 
slowed down accordingly. This results in abundant partly altered parent minerals such as 
feldspars and micas along with the formation of calcretes (Sunderam and Pareek, 1994). 

Doleritic rocks were observed at R3 exposure (see Fig. 1.1) and thin section of 
this doleritic rock under microscope (Plate 4.4) shows feldspar laths surrounded by altered 
augite (chloritised) ground mass. It indicates that the mineral alteration taking place here is 
may be due to the oxidised environment brought about by high temperature climate, and 
can be shown as 

Augite + + O 2 + H 2 O = Chlorite + Goethite + Ca^^ + ET 

The plagioclases present in this rock are mainly Ans? to Aneo as calculated from their 
extinction angle. Presence of some high Na-plagioclase (albite) is also reflected by albite 
twinning of thin lamellae type with low extinction angle. 

The weathered quartzite rock sample from lake bed shows a large deposition of 
iron oxide along the cracks of the rock (Plate 4.5). The presence of pyrite (FeS 2 ) in the 
form of small yellow colour tints on the surface of the rocks as identified in hand spedmen 


Plate 4.5 Quartzites at lake bed showing deposits of iron oxides 
ir them(450X, crossed nicol) ^ 

collected from lake bed, and the presence of Na2C03 in the lake brine suggest that the 
probable reaction taking place in the deposition of iron oxide is 

2 FeS 2 + O2 + H2O + Na2C03 = Na2S04 + Fe203.nH20 + CO2 + 4 S 

Here, the photosynthetical algae present in the environment is playing a major role by 
removing the CO2 from the system and thus putting the reaction in the forward direction. 


4.4.1 Methodology 

The sediments sampled in the field were analysed for grain size distribution 
through dry sieving. Prior to sieving the samples were digested in dilute HCl acid for the 
liberation of sand grains, and then, were washed in distilled water and properly dried. 100 
gms of these samples were then sieved for 20 minutes using mechanical shaker. For size 

analysis of these fiiable sands, a -^(j) (phi) sieve interval was used to obtain a reasonable 

accuracy. Individual size fraction of each samples thus obtained were weighed and noted 
in a tabular form for the required statistical calculation. 

4.4.2 Results and discussion 

Using Folk’s formula, the various size parameters, such as mean grain size and 
standard deviation have been calculated. The average values of statistical parameters of 12 
samples are listed in table 4 . 1 . 


Table 4.1: Statistical grain size parameters of different samples 
(J = Jhapok, M = Mendha) 











































The bivariate plots between mean size and standard deviation were generated (Fig. 
4.1) for the obtained data listed in table 4.1. Friedman (1967) has considered that plotting 
mean size Vs standard deviation provides an effective discrimination between river, dune 
and beach sands. This plot reveals that the samples from Mendha river channel and those 
from lake bed are forming two distinct clusters. Since the lake has no significant tidal or 
current action, this difference in grain size characteristics between the lake sediments and 
those of the river (Mendha) feeding it is either due to the different source of sediments or, 
largely, due to the chemical reaction of the same sediments in the lake water. The later one 
seems more likely case here in this lake environment, as the finer particles after taking rest 
in the lake shows progressively more “geochemical activity”. 

Size frequency curves (Fig. 4.2 a and b) have also been analysed for identifying 
sedimentary processes and depositional environments, as suggested by Visher (1969). The 


standard Deviation 


Lake bed sampies 


2.50 2.75 3.00 3.25 3.50 3.75 4.00 

Mean Size (phi) 


100 . 00 ^ 
O 2 


I 1 . 00 ^ 

I 0 . 10 ^ 

3 3 

O 71.- 

o T- w rt lo 

Depth = 0 to 30 cm 







^ -4 

1 20 - ! 

® 1.00 

1 ^ 


0 q-rp-fr *• 
o T” CM cn 'V in 

1 0.10 




1 = 30 to 60 cm 


Tpepth “ 60 to 90 cm 

o ■«- CM fo in 


2 3 4 5 


1 2 3 4 5 


1 2 3 4 5 



I = 


D* = 

I 0.10 

a - I 

1^: rh 

Depth = 90 to 120 cm 

LOO ^ 

/>?° “I 

/ o -\ 

'' E40 H 



Depth = 120 to 150 cm 

Fig. 4.2 (a) Log-normal and frequency curves of the sediment samples 
‘ from the Mendha river chaimel sediments 

cumulative frequency 

0 1 2 3 4 5 

Depth = 100 to 120 cm 

1 2 3 4 5 


Fig. 4.2 (b)Log-nonnal and frequency curves of the 
sediment samples from the lake bed at Jhapok 

entire grain size distribution of all the collected sediments consists of several log-normal 
poulations as inferred from the plots between phi, and cumulative frequency taken in 
logarithm scale. From the shape of the individual curves it is observed that the top three 
samples from Mendha river channel sediments are difiering from those present below it. It 
suggest that the top layer of the sediment up to 90 cm depth are wind transported, and the 
sediments present below to it are river transported. Hence, the river subsequently flowing 
into the lake during the monsoon must be carrying huge amount of wind contributed 

In the lake, the topmost layer (Fig. 4.2 b) has a different expression from the 
underlying sediments. When tried to correlate with the sediments of the Mendha river 
charmel, it was found that this topmost layer of the lake sediments is matching to certain 
extent with those collected at 120-150 cm depth from the river channel, except the 
subpopulations representing transportation through saltation are lacking in this lake 
sediment. However, all these plots representing lake sediments confirm that the large 
fraction of them were carried into the lake through suspension. 

The bar diagrams (Fig. 4.2 a and b) obtained by plotting phi Vs frequency also 
confirm the above discussion. In the Mendha channel sediments the wind derived top 
layers up to a depth of 90 cm mainly comprise of grains of the size range of 3.5 to 4.5 <j), 
while with depth the larger size fractions are dominating. However, all these plots for the 
Mendha channel samples are leptokurtic and shows normal distribution. In contrast, the 


lake sediments are of mixed population as indicated by their higher values of standard 
deviation with kurtosis and skewness being zero. 


X-Ray Diffraction (XRD) method is particularly useful for analysis of fine grained 
material which is difficult to study by other means. XRD provides the most efficient 
method for the determination of clay minerals in sediments, which may yield important 
information about provenance, weathering processes and depositional environment. 

4.4.1 Methodology 

The samples brought from the field were pretreated with 35% hydrogen peroxide 
to remove organic matter, cementing materials, etc. After this the samples were sieved to 
obtain the necessary size fraction between 270 and 325 A.S.T.M. sieve range for greater 
accuracy. The next step was to present the sample to the X-ray beam for which smear 
mount were prepared with distilled water on to a glass slide. This type of mount produces 
a partially oriented sample (Tucker, 1991). The smear mount is then inserted into the X- 
ray diffractometer ISO-DEBYEFLEX 1002 model of RICH SEIFERT & Co. at 20 MA/30 
KV using Cr-Ka radiation with monochromater. 

After the X-Ray chart of the individual sample between 6° and 50® of 20 value was 
obtained, peaks were identified in terms of 20 angle and converted to lattice spacing using 
conversion chart of 20 to angstrong (D-spacing). The problem creating ragged and 
overlapping peaks were deconvoluted using the method suggested by Tucker (1991). 


Then the D-spacing of each peak were matched with the table of consecutive lattice 
spacing for the common clays provided by Joint Committee on Powdered Diffraction 
Standard (JCPDS). 

As the intensity of the diffraction pattern of a mineral in a mixture is proportional 
to its concentration, it is possible to make rough estimation of the relative proportion of 
the minerals in a sample by measuring their relative peak heights or areas (Tucker, 1991). 
In the present case, the peak area of the individual minerals were calculated by multiplying 
the peak height above background by their width at half peak height. However, in case of 
overlapping peaks, which make the area calculation difBcult, height was used as a measure 
of relative proportion. The relative intensities of the minerals present were calculated by 
taking the intensity of quartz present in that particular sample as the standard. While in 
case of correlating the quartz quantity of different samples, further, the highest quartz 
peak among all the sample was taken as unity and accordingly the intensity of quartz 
present in other samples was normalised. Then, the peaks representing other minerals in 
the samples were renormalised, and their proportion in samples collected depth wise were 
graphically represented. 

4.4.2 Results and interpretation 

The XRD analysis (fig. 4.3) of sediments from the Mendha river bed and the soil 
sample near the rock exposure R1 (see Fig 1.1) shows the presence of illite and along with 
quartz and feldspar. The intensity of the illite peaks in Mendha river bed sediments 
indicate that the illite component in the sediment is not varying with depth. This is because 



the flat, low-lying, sandy river bed experiences runoff only during the monsoon; the 
infiltration of water after rain is at a maximum, and the subsurface drainage might be 
sluggish so that the soluble products released by hydrolysis reactions persist in the 
environmental water, thus inhibiting the further breakdown of the parent minerals. As the 
parent sediments are relatively rich in alkalis and alkaline earths and as the water from the 
nearby lulls are suggested to be charged with such ions the environment become distinctly 
alkaline. The water table is at a shallow depth (about 5m) as reported in open wells, and at 
some place above ground level forming brackish swamps, thus creating a strongly 
reducing environment. The ground water data of Yadav (1995) from various locations 
also indicate brackish environment. 

The XRD of lake sediment samples taken up to a depth of 1.6 m by augering near 
Jhapok (J in location map) shows that clay minerals present are mainly illite, chlorite, 
smectite, and their mixed layered clays. The XRD pattern of these sediments is shown 
depth wise in Fig. 4.4 and their relative proportion is represented graphically in Fig. 4.5. A 
careful scrutiny of Figs. 4.4 and 4.5 reveals that feldspar is absent at depths below 20 cm. 
Smectite appears at a depth of 20 cm. Also the mixed layer clay is c hanging from chlorite- 
montmorillonite type to montmorillonite-chlorite type indicating the increase in smectite 
proportion in the mixed layered clay after a depth of 20 cm. At a depth range of 80 to 100 
cm, the mixed layered clays type is appearing as mica-smectite type. Below a depth of 100 
cm smectite is completely absent from the scene. 

In contrast to the above pattern the XRD pattern of sediments from the lake bed at 
Gudha (Fig. 4.6) shows some kind of alternate zones. In the top 25 cm layer illite. 




intensity of minerai peaks intensity of minerai peaks 

Fig 4.5 Block rjiagram showing depth wise variation in concentration of 
different mineral 

160 J GUDHA 



chiorite-montmorillorate are present along with detrital quartz and feldspar. In the next 25 
cm depth the increase in the quartz content may be due to the precipitation of authigenic 
quartz. The simultaneous decrease in concentration of chlorite-montmorillonite and 
montmorillonite-chlorite mixed layer clay suggest that this layer has developed a low pH 
zone probably due to high organic activity. Thus the silica dissolved out from the top and 
bottom layer (higher pH zone) is precipitating here. The third layer of the depth range of 
50 to 75 cm is quite identical to the topmost layer (Fig. 4.5 and 4.6). The bottom-most 
layer again shows deposition of quartz and decrease in the quantity of mixed layered clay. 

The clay mineralogy of various section at Jhapok and Gudha site suggest a series 
of mineralogical transformations. The section at Jhapok being deeper shows that complete 
cycles on the basis of which two major zones of clay mineral assemblages can be 
recognised with a transitional layer in-between. At Gudha, only a part of the cycle is 
apparent partly because of shallow section and partly because of its location with respect 
to the lake. The description and details of the zones of clay mineral assemblages presented 

ZONEl (0 - 80 cm) 

• Quartz, orthoclase and plagioclase present in the topmost layer (0 to 20 cm) is 
considered to be transported. As discussed previously, illite and chlorite are forming in 
the source rocks and sediments outside the lake bed as alteration product, so their 
presence in this topmost layer is suggested to be detrital in nature. It is difficult to find 
out whether the chlorite-illite mixed layered clay present in this layer is detrital or 


authigenic. However, its origin due to leaching of pre-existing chlorite and mica can 
not be ruled out. 

• The decrease in plagioclase proportion and the absence of orthoclase in the second 
layer (20 to 40 cm) suggest that the from the lake water is replacing present in 
the crystal structure of these minerals and thus forcing them to go into solution. The 
increase in the illite proportion is the result of dissolution of feldspar, and the shifting 
of illite peak in XRD pattern is probably due to the absorption of in it’s stmcture. 
The chemical reactions suggested to be taking place is- 

Feldspar + BT = HAlSisOg + 

degraded silicate (illite) + cations (K^ + HCO^' + Si02= cation-Al-silicate (illite) + CO2 + H2O 

• The appearance of montmorillonite (smectite) in the second layer is due to the 
reaction of feldspar and ferromagnesian minerals in this water logged alkaline soil. The 
conversion of mixed layered chlorite-montmorillonite type to montmorillonite-chlorite 
type is obvious as smectite is precipitating. The decrease in quartz content and the 
removal of feldspar suggest the following reactions are taking place. 

SiOa + 2H2O = H4Si04 

Feldspar + 6IT + 2H4Si04 — > Smectite + 4H2O + Na"" + Ca'^ 

• The above trend is following into the third layer ( 40 to 60 cm) in an intense way. Here 
complete removal of feldspars is indicated by the absence of their peaks in the X-Ray 
chart. The quartz is going into the solution as in the previous case. The increase in the 


concentration of smectite in comparison to the above layer is taking place for the same 
reasons. Chlorite concentration is increasing probably due to the leaching of illite and 

• The transitional layer (80 to 100 cm) is designated so as the trend of the layers bellow 
is completely different from those above. Probably in this layer lots of organic (algal) 
activity taking place and thus the Eh value is sharply decreasing making 
montmorillonite equilibrium with the system. 

ZONE2 (im - 160 cm) 

This vertical zone below 100 cm depth shows a completely different pattern from 
the zone lying above it. Disappearance of smectite, slow increase in the illite content and 
precipitation of authigenic quartz are the major events noticed from the XRD pattern of 
sediments from this depth zone. 

• The first layer ( 100 to 120 cm) of this zone is experienced by the dissolution of 
smectite and chlorite as the high pH produced at this depth due to bacterial reduction 
of sulfate and possibly base exchange causing disequilibrium of clay minerals with the 
environment. The simultaneous precipitation of quartz and illite is obvious as the 
solution is getting saturated with respect to these minerals. 

• The mineralogjcal variation in the second layer(120 to 140 cm) of zone 2 suggest that 
due to the precipitation of authigenic quartz in the low pH region, the pH of the 
sediment water is increasing. This phenomenon is schematically shown in Fig. 4.6. 
With increasing pH montmorillonite is completely replaced by illite as follows : 


Smectite + ^ Mte + 

Chlorite is increasing at the expense of illite and hence changing the type of the mixed 
layered clay. The reaction can be shown as; 

Illite + MontmoriUonite-Chlorite + ->■ Chlorite + Illite-Montmoiillomte 


From all above observation it can be summarised that though the rocks and 
sediments are undergoing chemical weathering, but their soluble products are not removed 
to their best extent by the water, and subsequently the removed soluble constituents are 
reprecipitating on the surface of the rocks and sediments during the dry spell. 

Wind is a dominant agent responsible for most of the erosion and sediment 
transport work. Thus the loose products of the weathering and the reprecipitated soluble 
products are mainly transported to the lake bed by wind. The fluvial transport process is 
dominant particularly during the monsoon. The distinct variations in the size characteristic 
between the sediments in the lake bed and those in the Mendha river bed is due to the 
geochemical processes taking place in the lake. 

The weathering of sediments in lake bed is completely different from the 
weathering taking place in the adjacent plain or rock exposures. Inside the lake all kind of 
redox reactions are taking place at various depth causing mineral concentration as well as 
mineral transformation. The algae present in the lake water and sediments are governing 
the ongoing chemical reactions to a large extent. 




The work has been focussed at a rather complex geological phenomenon in a 
hypersaline lacustrine system. The Sambhar lake system, Rajasthan has been studied for its 
geomorphology, sedimentology, and catchment weathering. A veriety of techniques have 
been used for the study viz. remote sensing data alaysis coupled with field investigations, 
sedimentological analysis of river and lake sediments, clay mineralogical analysis df 
catchment rocks, river and lake sediments and also limited microscopic analysis of rock 
thin sections. Based on the results of these analyses, a conceptual model for the physical 
and chemical evolution has been evolved. The details of the model are presented next. 


The regional geomorphological features have been studied using the satellite 
remote sensing data. The geomorphological and tectinic features present in the area as 
observed in the field and on satellite imageries (Chapter 3) suggest the sequence of events 
that might have led to the formation of the Sambhar lake (Fig. 5.1). The first post-orogeny 
movement might have caused the disruption in the Aravalli range (Dassarma, 1988) and 
during this phase, the Block B was shifted towards west along two strike slip faults 
demarcating its boundary in the North and South respectively. Its continuation in the east 
could not be ascertained as the satellite data of this area was not available for the present 

study. The truncated hills, bending and rupturing of Aravalli hills (plate B.l & fig. 3.5) in 
this area supports this interpretation. After this, the post orogeny movement during Plio- 
Pliestocene period (Sen & Sen, 1983) might have caused the upliftment of the blocks, on 
either side of Block B, along reactivated strike slip faults. This might have ripped open the 
linear depression forming the Sambhar lake. 

Sambhar lake, a quaternary sedimentation basin as described by Roy & Sen 
(1983), has probably imdergone impulsive movements as the presence of faults (Fig. 3.5) 
in block A suggest. However, this quaternary sedimenatary basin was probably getting the 
sediments from the river flowing along the E-W trending lineaments (Fig. 3.4). Probably 
due to filling up of this basin or due to some tectonic movements in the upstream reaches, 
the river channels later changed their course and left their impression in the form of 

Further, the basin has undergone extensive wind deflation. The high wind flowing 
from the west along the gap in the Aravallis initiated the deflation process excavating a 
hollow in this quaternary sand. The presence of huge sand dimes and sand mounds along 
the eastern periphery of the lake supports this intrpretation. The wind action which 
followed can be described using a geographical model proposed by Greelet & Iversen 
(1987). According to this model, lee waves often form where mountain ranges are located 
upwind from fiat plains. When air flows over the crest of this range of mountains, it must 
rise in order to do so. In turn, it descends when traveling over the leeward slope. The 
momentum gained by the air as it travels down, the slope carries it past an equilibrium 


condition, forcing it to ‘bounce’ back up again. The momentum gained the upward motion 
again carries it past an equilibrium condition and the air starts downward a second time. 

The standing wave pattern thus formed can exist for several oscillations as seen in Fig. 

5. ID. As a result, this wind blows out the soft materials i.e. sand, if any, present on the 
surface towards the lee side. The present-day Sambhar lake is therefore an outcome of 
tectonic and extensine aeolian activity in the region. 


The available data on the brine chemistry of Sambhar and the sediment-water 
interaction phenomenon presented in this study (Chapter 4) provides valuable insight into 
the chemical evolution of the Sambhar lake. The product of catchment weathering, 
together with a certain amount of unweathered rock material, is carried by wind and river 
into the Sambhar lake. During transportation, the rock and mineral particle largely 
undergo mechanical weathering, and little chemical weathering of silicates takes place to a 
far greater extent in the lake than on the adjoining plains because large amount of 
chemically active water is available. All the physiochemical processes occurring in this 
hypersaline lake environment can be termed as “halmirolysis”, and is responsible for the 
existing Na-Co 3 -So 4 -Cl type brine in the lake (Bhattacharya et al, 1982; Raymahashay, 


Fig. 5.2 shows the various steps of lake halmirolysis as inferred from sediment ! 


mineralogy. When the unaltered detrital sediments comes in contact with lake water, ; 






Fig. 5.2 Schematic diagram showing geochemical evolution of the 
Sambhar lake brine 


halmirolysis starts. As a result, the soluble chemical constituents like Na"^, Mg^, K'", Ca"^, 
etc. in the sand and silt sized mineral fractions get mobilised but they still remain in the 
interstitial solution making the environment more alkaline. The already existing detrital 
clays in sediments i.e. illite and chlorite absorbs these ions to produce new minerals or 
change into some mixed layered type clays. 

The smectite in the sediments is transformed stepwise into chlorite as observed in 
the case of Jhapak sediments, suggesting increased salinity, as a thumb rule with increasing 
salinity almost all the clay species become decreased with exchangeable cations. These 
interstitial solution get enriched with respect to Na"^ and Ca^^, the former being dominant. 
Due to summer heat or mechanical squeezing the interstitial water moves from deeper 
horizon either vertically or to some extent laterally (up dip) into higher formations in 
which the chemical equilibrium may be completely different from that of the underlying 

Here, considering the possibility of the operation of “selective chemical filtering”, 
it can be explained that the passage of interstitial water heavily charged with organic and 
inorganic solutions and collides, through a porous membrane is likely to cause filtering of 
large molecules and ions of opposite charges to that of the membrane. In as much as 
quartz in the sediments carrying no charge, the separation is mechanical, i.e. differential 
capilarity. As the clays are present to a great extent in the sediments this filtering is also 
chemical. This is obviously taking place as montmorillonite (smectite) which has strong 
negative charges is present in the sediment. At first the passage of the negatively charged 


ions are mechanically restricted and then the corresponding cations (Ca^"^, Na^ & 
K‘") get trapped. 

In the meantime, calcite precipitates out with evaporation which quantitatively 
removes Ca^^ ions from the sediments solution. The calcite precipitation is well reported 
by XRD of xjntreated sediment samples in the lake bed at Jhapak. After the precipitation of 
calcite no pure Ca-mineral precipitate. However, subsequently Mg-minerals like Mg- 
smectite precipitate out which is well observed in the near surface (zone 1) zone of Jhapak 
sediments. In this zone the presence of large quantity of Ulite strongly suggests that 
fixation of K'^ in three layered expansible clay i.e. smectite, is too strong to be exchanged 
by ordinary cation exchange capacity reaction. 

After all these conversion, fixation and transformation processes the final 
interstitial water which apparently move upward due to overburden or due to evaporation 
from the final composition either of Mg^'^-Na* -SO4-CI type or Na-C03-S04-Cl type, 
depending upon the Mg^^ / carbonate alkalinity. The Sambhar lake brine has been reported 
to be of Na-Co3-So4-Cl type (Bhattacharya et al, 1982; Raymahashay, 1996). The model 
outlined above therefore still does not explain the presence of 804*^ and Cl' in the lake 
brine which further probing of sediments and integrating the detailed water chemistry data 
into the model. 



Sambhar lake is a complex geological and geochemical system and the work 
presented in this thesis is still inadequate to explain its evolution completely. The following 
aspects need further attention to unravel the complexities of this very interesting area ; 

1 . A modified algorithm may be needed to distinguish vegetation in the desert area in the 
processed satellite data, which in turn would have helped in locating the lineaments. 

2. The sharp difference in reflectance of lake water jfrom different parts of the lake seems 
very interesting and it must be related to dififemce in salinity level, type of algal growth 
and other biogeochemical parameters which needs further investigation. 

3. Laboratory experiment is needed to find out the change in size parameter of the 
provenance sediments, when put in a solution similar to lake, which suppose to help in 
justifying the discrimination in bivariate plot. 

4. The pore fluid chemistry of lake sediment samples may provide vital clues to the 
geocjemical cycle responsible for the origin of hypersalinity in the lake. 



Anhert, F. (1970) Functional relationship between denudation, releif and uplift in large 
mid-latitude drainage basins, Am. Jour. Sci. v. 268, No.3., pp.253-263. 

Bhattacharya, B. N., Gupta, S.K., Rishi, B.P., Mitra, A. K. & Kashyap, V. K. (1975), 
Seasonal Variation in composition of Sambhar lake brine and possible origin of its 
salinity, Proc. Workshop on the problem of deserts in India, GSI, Jaipur. 

Blatt, H., Middleton, G. V. and Murray, R. (1980) Origin of sedimentary rocks, 2nd ed. 
Englewood Cliffs, NJ: Prentice Hall. 

Bloom, A. L. (1969), The surface of the earth, Prentice-Hall, Inc., Englewood Cliffs, New 

Dassarma, D. C. (1988) Post orogenic deformation of the Pre- Cambrian Crust In North- 
Eastern Rajasthan, In; Precambrian of the AravalU Mountain, Rajasthan, India; 
geol. Soc. India; Bangalore; Memoir 7, pp. 109-120. 

Eugster, H.P. & Hardic, L. A., (1978) Saline lakes; In: Lake Chemistry Geology Physics, 
Spiringer verleg. 

Godbole, N. N. (1952), “The salinity of the Sambhar lake”, BuU. Nat. Inst. Sci. Ind., 1, 

Gopal, B. and Sharma, K. P. (1994) Sambhar lake Rajasthan (Ramsar site of India), WWF 
publication. 37p 

Greelet, R. and Iversion, J. D. (1987) Wind as a geological process, Cambridge planetary 
Sc. series, Cambridge University Press. 

Gupta, R. P. (1991), Remote Sensing Geology, Spinger-verlag, Berlin Heidelberg 

Jensen, J.R. (1986), Introducing Digital Image Processing, Pretince-Hall, New Jersey. 

Lillesand, Thomas M. and Kiefeer, Ralph. W. (1994) Remote Sensing and Image 
Interpretation, - 3rd ed. John Wiley & Sons. 

Mokhopadhya, D. and Dsagupta, S., (1978) Delhi- preDelhi relations near Bandar, Central 
Rajasthan, Indian Jour, of Earth Sci., v.7, pp.64-75 . 

Ray, Terril W. and Murray, Bruce C. (1996), Nonlinear Spectral Mixing in Desert 
vegetation. Remote Sens. Environ, vol.55 pp.59-64. 

Raymahashay, B.C. (1996), Geocemistry for Hydrologists, Allied Publishers Limited. 

Rees, C.C. (1968) Introduction to paleolimnology; Development in Sedimentology, 
Elisewer publishing company. 

Roy, A. B. (1988) Stratigraphy and Tectonic framework of The Aravalli Mountain 
Range; In: Precambrian of Aravalli Mountain, Rajasthan, India; Geol. Soc. Of 
India, Bangalore; Memoir 7, pp. 3-32. 

Roy, A. B. and Das, A. R. (1985) A study of the time relations between movements. 
Metamorphism and Granite emplacement in the middle Proterozoic Delhi 
Supergroup Rocks of Rajasthan, Jour. Geol. Soc. India, v. 26 pp. 229-236. 

Roy, J. N. and Sen, S. (1983) Post neogen tectonism along the Aavalli range, Rajasthan, 
India, v24, pp. 229-236. 

Sen, D. and Sen, S. (1983) Post neogen tectonis along Aravalli range, Rajasthan, India, 
Tectono Physics, V. 93, pp. 75-98 

Sharma, Ram S.(1988), Pattern of Metamorphism in the Precambrian rocks of the Aravalli 
mountain, Rajasthan, India; Geol. Soc. of India, Bangalore, Memoire 7, pp. 33-77. 


Singh, S. P. (1988) Stratigraphy and Sedimentation pattern in the proterozoic Delhi 
Supergroup, North western India, In: Pre Cambrian of Aravalli Mountain 
Rajasthan, India, Geol. Soc.of India, Memoir 7, pp 193-206 

Sundaram, R. M. and Pareek, S. (1995) Quaternary facies and paleoenvironment in north 
and east of Sambhar lake, Rajasthan. Jour. Geol. Soceity of India, v.46, pp. 385- 

Tucker, M. (1991) Techniques in sedimentology, Blackwell Scientific Publications. 

Yadav, D.N. (1995) Chemical and isotopic studies in coastal area of the Arabian sea and 
saline lakes of Rajasthan (India). Unpublished Ph. D. thesis, M. S. University,